Neural Network based pulse shape analysis with the Belle ...
Transcript of Neural Network based pulse shape analysis with the Belle ...
Neural Network based pulse shape analysis with the Belle II electromagnetic calorimeter5th of November 2020
Women Physicists' Conference
Stella Wermuth
Belle II overview
• SuperKEKB 𝑒+𝑒− collider in Japan
• Collision energy: 10.58 GeV
• Main goals of Belle II:• Precision measurements
• Rare/forbidden processes
• Beyond Standard Model search
https://www.belle2.org/project/super_kekb_and_belle_ii
𝑒−
𝑒+
Stella Wermuth - Belle II Pulse Shape Analysis 2
• ~9000 scintillator crystals CsI(Tl)
• Main task: • Measuring electromagnetic energy
• Reconstructing neutral particles
• Particle identification using shower shapes
Electromagnetic Calorimeter (ECL)
https://www.belle2.org/project/super_kekb_and_belle_ii
Stella Wermuth - Belle II Pulse Shape Analysis 3
Analysis: 𝑒+𝑒− → 𝜋+𝜋−𝛾
Need good particle identification!
Belle II, https://docs.belle2.org/record/1124/files/BELLE2-NOTE-PL-2018-032.pdf
Stella Wermuth - Belle II Pulse Shape Analysis 4
Analysis: 𝑒+𝑒− → 𝜋+𝜋−𝛾Belle II, https://docs.belle2.org/record/1124/files/BELLE2-NOTE-PL-2018-032.pdf A. Keshavarzi, D. Nomura, T. Teubner arXiv:1911.00367 [hep-ph]
Theory Experiment
Improve precision of 𝑒+𝑒− → 𝜋+𝜋−
measurement
Reduce uncertainty of SM-calculation of the anomalous
magnetic momentum of the muon
Stella Wermuth - Belle II Pulse Shape Analysis 5
What happens in the ECL?
AD
C c
ou
nts
digitized pulse shape
6
time
= ECL crystal
= ECL crystal where energy is deposited
Stella Wermuth - Belle II Pulse Shape Analysis 6
What happens in the ECL?
AD
C c
ou
nts
digitized pulse shape
7
time
FPGA• Fit pulse shape• Reconstruct Etotal
• Optimized for photons
FPGA Fit
Stella Wermuth - Belle II Pulse Shape Analysis 7
Particle Identification with the ECL
AD
C c
ou
nts
time
AD
C c
ou
nts
time
8
fast „hadron“ scintillation component
Stella Wermuth - Belle II Pulse Shape Analysis 8
Pulse Shape Discrimination• Multi-template fit (photon and hadron template)
• Offline
• Reconstruct Etotal, Ehadron
Hadron Intensity = EhadEtot
Savino Longo, https://dspace.library.uvic.ca/handle/1828/11301
Stella Wermuth - Belle II Pulse Shape Analysis 9
Keywords to remember:
• Pulse shape:• 31 ADC points
• Difference between photon-like and hadron-like pulse shapes
• Default methods:• FPGA: Etotal, optimized for photons (online)
• Multi-template: Etotal, Ehadron (offline)
• Hadron Intensity = EhadEtot
Stella Wermuth - Belle II Pulse Shape Analysis 10
Neural Network approach
Multi-template fit
Neural Network
Etot, Ehad
What can we gain by using a Neural Network instead of a multi-template fit? (speed, precision, robustness)
?
AD
C c
ou
nts
time
Stella Wermuth - Belle II Pulse Shape Analysis 11
Neural Network structure
• Fully connected NN
• 2 hidden layers (256 neurons each)
• 31 inputs (normalized ADC counts)
• 2 outputs (Etot, Ehad)
Stella Wermuth - Belle II Pulse Shape Analysis 12
Neural Network inputParticle:
Photonsas proxy for EM interactions
Pionsas proxy for hadronic interactions
Stella Wermuth - Belle II Pulse Shape Analysis 13
Neural Network inputParticle:
Stability tests:Baseline
fluctuationsDifferent particle
test data
AD
C c
ou
nts
time
Photonsas proxy for EM interactions
Pionsas proxy for hadronic interactions
Stella Wermuth - Belle II Pulse Shape Analysis 14
Neural Network inputParticle:
Stability tests:Baseline
fluctuationsDifferent particle
test data
AD
C c
ou
nts
time
Photonsas proxy for EM interactions
Pionsas proxy for hadronic interactions
Stella Wermuth - Belle II Pulse Shape Analysis 15
Neural Network input photons
• Trained with ~400 000 pulse shapes
• Tested with ~80 000 pulse shapes
Stella Wermuth - Belle II Pulse Shape Analysis 16
Neural Network Etotal prediction
linear response
Stella Wermuth - Belle II Pulse Shape Analysis 17
Neural Network Etotal resolutionEtot resolutiondifference of true value and prediction
Stella Wermuth - Belle II Pulse Shape Analysis 18
energy resolution improved
Neural Network inputParticle:
Stability tests:Baseline
fluctuationsDifferent particle
test data
AD
C c
ou
nts
time
Photonsas proxy for EM interactions
Pionsas proxy for hadronic interactions
Stella Wermuth - Belle II Pulse Shape Analysis 19
Baseline fluctuations: fittype (FT)
• Assigned by multi-template fit• Fittype = 0 good• Fittype = 1 additional peak• Fittype = -1 bad
Stella Wermuth - Belle II Pulse Shape Analysis 20
Baseline fluctuations: fittype (FT)
Stella Wermuth - Belle II Pulse Shape Analysis 21
Neural Network inputParticle:
Stability tests:Baseline
fluctuationsDifferent particle
test data
AD
C c
ou
nts
time
Photonsas proxy for EM interactions
Pionsas proxy for hadronic interactions
Stella Wermuth - Belle II Pulse Shape Analysis 22
Neural Network input pions
• Trained with ~328 000 pulse shapes
• Tested with ~82 000 pulse shapes
Stella Wermuth - Belle II Pulse Shape Analysis 23
Neural Network Etot & Ehad resolutionEtot resolution Ehad resolution
hadron resolution improved for photon-like pulse-shapes
Stella Wermuth - Belle II Pulse Shape Analysis 24
Neural Network inputParticle:
Stability tests:Baseline
fluctuationsDifferent particle
test data
AD
C c
ou
nts
time
Photonsas proxy for EM interactions
Pionsas proxy for hadronic interactions
Stella Wermuth - Belle II Pulse Shape Analysis 25
time
Train with pions, infer on photonsEtot resolution Ehad resolution
total and hadron energy resolution improved up to 200 MeV
Stella Wermuth - Belle II Pulse Shape Analysis 26
What is next?
Running Neural Network
Data & Analysis
• Move towards data• Look at 𝜋𝜋𝛾 analysis• See if NN can improve 𝜋/𝜇-separation• Reduce 𝜇-background• Reduce systematic uncertainty
Belle II, https://docs.belle2.org/record/1124/files/BELLE2-NOTE-PL-2018-032.pdf
Stella Wermuth - Belle II Pulse Shape Analysis 27
Summary
• Running Neural Network for pulse shape analysis
• Performance compared with multi-template and FPGA fit
• Results:• Improved Etot and Ehad-resolution (for pions up to 200 MeV)
• Stable with respect towards fittype
• Interesting application: 𝑒+𝑒− → 𝜋+𝜋−𝛾
• Implement ML on FPGA (new PhD project starting soon)
Stella Wermuth - Belle II Pulse Shape Analysis 28
Back Up.
Traditional Particle Identification
• Measure track → momentum (CDC)
• Measure velocity (TOP)
• Combine information for mass
• 𝜋/𝜇-separation hard, tracks too close
Stella Wermuth - Belle II Pulse Shape Analysis 30
• High precision measurement of the anomalous magnetic momentum of the muon
• 3.7𝜎 discrepancy (arXiv:2006.04822, 2020 paper)
• Theory limited by 𝑒+𝑒− → 𝜋+𝜋−
Excursion: g-2 muon
A. Keshavarzi, D. Nomura, T. Teubner arXiv:1911.00367 [hep-ph]
Theory Experiment
Stella Wermuth - Belle II Pulse Shape Analysis 31
Excursion: theory of g-2
QCD:Needs to be
measured
A. Hoecker and W. J. Marciano, 2013 „The Muon Anomalous Magnetic Moment“
Stella Wermuth - Belle II Pulse Shape Analysis 32
Excursion: theory of g-2
A. Hoecker and W. J. Marciano, 2013 „The Muon Anomalous Magnetic Moment“
QED Universality
Improve precision of g-2 theory by improving precision of the
measurement of this process
T. Aoyama et al. arXiv:2006.04822 [hep-ph]
Stella Wermuth - Belle II Pulse Shape Analysis 33
Analysis: 𝑒+𝑒− → 𝜋+𝜋−𝛾
arXiv:2006.04822 [hep-ph]
Stella Wermuth - Belle II Pulse Shape Analysis 34
FPGA Fit within the ECLA. Sibidanov, https://docs.belle2.org/record/800/files/BELLE2-TALK-CONF-2018-019.pdf
• Two photodiodes• Pre-amplifier (integrate and shape the signal)• Sum• ShaperDSP (shaping amplifier, tail subtraction)• Digitizer (into 31 ADC points)• FPGA (template fit, photon)
Savino Longo, https://dspace.library.uvic.ca/handle/1828/11301 - Chapter 6.1
Stella Wermuth - Belle II Pulse Shape Analysis 35
• All pulse shapes with energy above 50 MeV are stored for offline analysis
• Photon template: • pure photon scintillation component
emission• Use Bhabha scattering events for
computing the templates
• Hadron template• pure hadron scintillation component
emission• Compute with input from signal chain
response and testbeam study
• Need to compute signal chain output for calibration (11 parameters for template for each crystal)
Multi-template fit
Savino Longo, https://dspace.library.uvic.ca/handle/1828/11301 - Chapter 6.3
Savino Longo, https://dspace.library.uvic.ca/handle/1828/11301
Stella Wermuth - Belle II Pulse Shape Analysis 36
Cluster resolution
Stella Wermuth - Belle II Pulse Shape Analysis 37
Neural Network hadron intensity
Stella Wermuth - Belle II Pulse Shape Analysis 38
Beam background Trained with run 3363, infered on run 3363 and 5649 separatly
Stella Wermuth - Belle II Pulse Shape Analysis 39